Aladdin - Scala Bugtracking
[#133] project: compiler priority: high category: bug
submitter assigned to status date submitted
Matthias Burak fixed 2003-09-11 15:33:33.0
subject ClassCastExceptions at runtime
code 
class Expr;
case class One(xs: List[Expr]) extends Expr;
case class Two() extends Expr;
object Foo with Executable {
  def test(xs: List[Expr]) = xs match {
    case List(a@(Two()*)) => System.out.println("a = " + a);
  }
  test(List(Two(),Two(),Two(),Two()));
  test(List(Two(),Two(),Two()));
  test(List(Two(),Two()));
  test(List(Two()));
  test(List());
}
what happened 
Compiles, but at runtime it yields the following exception:

Exception in thread "main" java.lang.ExceptionInInitializerError
        at Foo.main(Foo.scala)
Caused by: java.lang.ClassCastException
        at Foo$.test(Foo.scala:1)
        at Foo$.(Foo.scala:8)
        at Foo$.(Foo.scala)
        ... 1 more
what expected Compiles and runs without errors.
[back to overview]
Changes of this bug report
Burak  edited on  2003-09-23 18:25:25.0
change priority to high
Burak  edited on  2003-09-24 14:09:25.0
this one is not reproducible anymore. Very likely that the fixing of previous errors fixed also this one, I do not know. Syntax is now 
 List(a@(Two()*))
.
Matthias  edited on  2003-10-01 00:42:54.0
I just re-opened the bug again, because it seems to be not fully fixed. A small variation of the program above yields the same effect. Here's the program: 
case class One();
object Foo with Executable {
  def test(xs: List[Any]) = xs match {
    case List(x@(One()*), y@(One())) => System.out.println("case");
    case _ => System.out.println("default");
  }
}
The runtime exception is: 
Exception in thread "main" java.lang.ExceptionInInitializerError
        at Foo.main(D.scala)
Caused by: java.lang.ClassCastException
        at Foo$.test(D.scala:1)
        at Foo$.(D.scala:7)
        at Foo$.(D.scala)
        ... 1 more
Burak  edited on  2003-10-06 19:47:38.0
the bug was a wrong condition in Tracer2Scala.
the distinction between bindings x@_ and x@_* is correct now.